Based on our experience with acute myelogenous leukemia (AML), we have hypothesized that blast crisis cells from chronic myelogenous leukemia (CML) patients should display aberrant surface immunophenotypes. Initial studies have borne this out. We now intend to investigate the clinically relevant implications of these findings. We will investigate the prognostic significance of reappearance of blast crisis aberrant phenotypes in second chronic phase, and of their reappearance in remission after autologous or allogeneic marrow transplant. We will directly immunophenotype CML cells from patients in first chronic phase, in accelerated phase, and in complete cytogenetic remission and correlate the frequency of aberrant cells with disease progression. We will also correlate the frequency of aberrant cells in first chronic phase with existing prognostic models for CML. Finally, we will purge the marrow of aberrant CML cells and purify diploid hematopoietic stem cells for autologous transplant. We will examine the efficiency of removing t(9;22) CML cells using fluorescent-in situ hybridization (FISH), and examine the purified cells for bcr-abl transcripts using reverse-transcriptase polymerase chain reaction (RT-PCR) and the self-sustaining sequence replication reaction (3SR). Cells will be expanded in vitro using recombinant hematopoietic growth factors, and reanalyzed by these techniques. Purified stem cells will also be examined in vivo in a SCID- hu mouse model to evaluate their potential to provide normal diploid hematopoiesis and the absence of engrafting CML cells.